SandLogicTechnologies/Hermes-4-14B-GGUF

Quantized Hermes-4-14B Model

This repository provides a quantized GGUF version of the Hermes-4-14B model. The 4-bit and 5-bit quantized variants retains the model’s strengths in advanced reasoning tasks while reducing memory and compute requirements ideal for efficient inference on resource-constrained devices.

Model Overview

• Original Model: Hermes-4-14B
• Quantized Version:
  • Q4_K_M (4-bit quantization)
  • Q5_K_M (5-bit quantization)
• Architecture: Decoder-only transformer
• Base Model: Qwen3-14B-Base
• Modalities: Text only
• Developer: Nous Research
• License: Apache 2.0 License
• Language: English

Quantization Details

Q4_K_M Version

• Approx. ~69% size reduction
• Lower memory footprint (~9 GB)
• Slight performance degradation in complex reasoning scenarios

Q5_K_M Version

• Approx. ~64% size reduction
• Lower memory footprint (~10.5 GB)
• Better performance retention, recommended when quality is a priority

Key Features

• Reasoning that is top quality, expressive, improves math, code, STEM, logic, and even creative writing and subjective responses.
• Instruction-following model optimized for multi-turn scientific question answering
• Schema adherence & structured outputs: trained to produce valid JSON for given schemas and to repair malformed objects.
• Much easier to steer and align: extreme improvements on steerability, especially on reduced refusal rates

Usage Example

Text Inference:

./llama-cli -hf NousResearch/Hermes-4-14B-Q4_k_m.GGUF -p "Explain the Fourier Transform in simple terms"

Recommended Use Cases

• Scientific reasoning & STEM domains: tasks requiring step-by-step logical reasoning, clean structure.

• Coding & software-related tasks: code generation, explanation, debugging.

• Chatbots/Assistants: where reasoning transparency is important (showing chain of thought).

• Low-resource deployment / edge inference: use quantized variants.

Acknowledgments

These quantized models are based on the original work by the NousResearch development team.

Special thanks to:

• The NousResearch team for developing and releasing the Hermes-4-14B model.

• Georgi Gerganov and the entire llama.cpp open-source community for enabling efficient model quantization and inference via the GGUF format.

---

Contact

For any inquiries or support, please contact us at support@sandlogic.com or visit our Website.